Real-Time Upserts In Apache Pinot And Startree Cloud
StarTree, Thursday, November 30,2023
One of the most critical aspects of real-time analytics is the ability to derive timely and accurate insights from data in motion. Companies of all sizes are constantly generating data - whether that be new events (insert) such as clickstream, logs, and metrics, or updates to existing events (upsert) such as customer/employee profile changes.
The ability to ingest these upsert events in a real-time analytics database is essential for generating the right insights. Real-time upserts ensure the query results are accurate and reflect the latest changes happening to your business.
For example, Uber's financial intelligence and AdTech insights depend on accurate metrics generated from constantly updating events (like driver and rider location). Similarly, various risk and fraud detection use cases depend on the ability to derive instant and accurate insights from customer data managed by an Online Transaction Processing (OLTP) system (like MySQL, Cassandra, or MongoDB) that is constantly getting updated.